Tracking of shape-changing colloids via adaptive image reconstruction

Particle tracking methods are essential tools for probing the structure, dynamics, and interactions in colloidal systems. However, conventional methods often introduce systematic bias when particle images overlap. Image reconstruction algorithms can overcome these limitations, but they typically cannot handle in-situ changes in size and morphology-crucial aspects for stimuli-responsive colloids such as thermoresponsive microgels. Here, we introduce an adaptive full-image reconstruction method for radially symmetric particles in the micron-to-submicron size range that iteratively learns particle size and optical contrast directly from microscopy images, enabling simultaneous extraction of precise positions and morphologies. We apply this method to hybrid microgels loaded with gold nanoparticles, obtaining their position and shape.